Telecommunications companies process a variety of data records. One example of such data records are Call Detail Records (CDRs), which consist of data produced by telephone equipment and other communication devices, documenting the details of a call or connection, such as time, duration, completion status, source and destination.
A CDR is a metadata record for a telephone exchange or another communication transaction. A CDR typically indicates a start time, duration, completion status, and a source/destination pair of the corresponding communication, such as a voice call, a short-message services (SMS) message or another data exchange Calls between two cellular telephones, for example, would register the source/destination pair and cell tower in use, but data packets would only record the source and cell tower employed because a destination is not involved. Other relevant attributes of a CDR often include equipment identification, a unique sequence number from the telecommunications provider, a reference to beginning/ending cell towers, routing information, call type (e.g., voice or SMS), mobile telecommunication technology used, and any fault conditions during the call.
CDRs serve several purposes for telecommunications companies and other stakeholders. In law enforcement, for example, CDRs provide information that helps to identify suspects and other persons of interest. In addition, CDRs often reveal details about an individual, such as relationships with associates, communication and behavior patterns, and location information that can identify the whereabouts of an individual during the entirety of a call. CDRs are also used for billing, customer experience management, marketing planning and population movement patterns.
A need exists for a computational system for processing and organizing voluminous data record workloads, such as CDR workloads, that allow for the execution of real-time queries over pre-processed and indexed data records. A further need exists for improved techniques for processing and organizing voluminous data record workloads, allowing for the execution of real-time queries over indexed data records.